What sank Sendo's Stinker?
Back to the chalkboard for Microsoft phones
Analysis After you shake hands with Microsoft, goes the saying, count your fingers. Sendo's decision to abandon Microsoft's cellphone platform for the emerging Nokia standard, Series 60, gives the British company a new lease of life. But it's a crushing blow to Microsoft, which needed every OEM it could find for its Smartphone 2002 platform, and must now go it alone with an unproven and risky manufacturing model.
You never want to say never with Microsoft - the company has $40 billion in the bank and determination to succeed - but it has a mountain to climb if it's to restore its credibility. For four years the company has been talking about its cellphone plans, promising that its value proposition (commoditized hardware, enterprise connectivity) would prove irresistible, but it's largely friendless and getting lonelier by the day.
The most obvious option - the acquisition of a proven smartphone operating system and hardware design team - doesn't exist. There's only one candidate, SymbianOS, and that's not for sale. Dozens of other operating systems exist that could potentially fit the bill, but mature as they might be, they don't have the connectivity stacks - SyncML, Bluetooth - or the applications (messaging, Java) tailored to a communications device.
But software is only part of the problem - the hardware issue is much more serious. Nobody wants to play with Microsoft. Redmond has tried wooing the big manufacturers - with Samsung as the only success - and worked particularly hard with the smaller vendors - who Bill Gates unfortunately described as "the seven dwarves" in his famous cellphone strategy memos. This isn't the PC business, where when Bill calls a party, everyone is expected to attend. Only Sony is a significant Windows licensee, but Nobuyki Idei's antipathy towards Microsoft is legendary: he calls it an "OS dinosaur" that's unable to adapt to new business models.
Why doesn't anyone want to play? Short-term tech practicalities don't inspire confidence. Here are three: Redmond has at best an annual product release cycle, it isn't used to producing software that integrators can work with, and it has a poor reputation for reliability. Phone manufacturers - and Sendo told me half of the software in the aborted Z100 was their own - need access to the source and they need problems fixed fast. Today's phones have a brief TTL (time to live) on the market, and delays or a single recall can wipe out the profit on a particular model.
Longer-term, manufacturers don't want to be squeezed into a PC business model where no one other than Dell makes any money. Now you might think the phone business is going to become commoditized, and consider that a historical inevitability. But commoditization takes place in mature markets (the PC hasn't essentially changed in twenty years) where there's a minimal integration or testing phase, and minimal differentiation between products. Integration to a PC OEM means strapping on an anti-static belt, picking up a screwdriver, and a few minutes later you have a product (which looks like all the other products).
Integration in the phone business means network testing and gaining regulatory approval, which takes months; and style is a crucial differentiator: these are fashion items.
Microsoft must now "learn" this expertise itself: the answer we've known for some time, is that it designs the phone itself and gets a contract manufacturer (HTC, Sentel) to produce it. In which case Microsoft merely inherits the problems experienced by Sendo.
Here's what I think will happen. Expect to hear a lot fairly soon from Redmond about how the operating system doesn't matter. Expect to hear the phrase ".NET phones", and much chastened talk about being older and wiser, which you can safely ignore: it's a sure sign they're about to repeat some familiar mistakes.
Microsoft needs a story in which all today's smartphone and telephony standards are rendered passé, and the generic marketing fog of .NET fits the bill perfectly. (Anyone remember IBM's Systems Applications Architecture?)
But by the time .NET phones are ready appear - we're talking two to three years - smartphones won't look anything like what we expect today. Mobile devices will be all shapes and sizes, and just happen to have some connectivity built in. Nokia gaves us a glimpse of this on Monday: it looks like this.
What sank Sendo's Stinker will continue to send future Microsoft phones to a watery grave: unless Redmond can come up with some very un-Redmond-like thinking ®